OECD calls for broadband cost-benefit analysis

OECD calls for broadband cost-benefit analysis

Summary: The Organisation for Economic Co-operation and Development (OECD) has warned that governments need to weigh up investment in telecommunications infrastructure, which the opposition has used to pressure the Federal Government to conduct a cost-benefit analysis on its $43 billion NBN proposal.


The Organisation for Economic Co-operation and Development (OECD) has warned that governments need to weigh up investment in telecommunications infrastructure, which the opposition has used to pressure the Federal Government to conduct a cost-benefit analysis on its $43 billion NBN proposal.

"Policy makers need to evaluate the costs and benefits of any public investment in telecommunications infrastructure and select projects which can stimulate current demand, but simultaneously expand the productive capacity of the economy in the longer term," said the France-based OECD in a briefing pack released this week that detailed national broadband research.

The comments were immediately seized on by Shadow Communications Minister Nick Minchin as evidence for his claim that the government should carry out a comprehensive cost-benefit analysis on the NBN project before its commencement.

Communications Minister Stephen Conroy has denied the need for such a study and has insisted Minchin's stalling tactics were threatening the likely benefits of the NBN.

A spokesperson for Conroy today declined to directly respond to Minchin's comments or the OECD report, instead pointing to comments Conroy made last week in a speech.

Conroy said the NBN would cost less than the $43 billion proposal on the basis that current network owners, such as Telstra, could "vend-in" to the NBN.

"We expect that there will be substantial private sector interest in this network. This includes the possibility that companies will want to vend-in existing assets that can support the National Broadband Network for equity or some other financial arrangements," said Conroy last week.

"We expect that the actual cost to be significantly lower than $43 billion for a number of reasons, including the substantial contingency intentionally built into the estimate," he said.

But even if Minchin does force the government to submit the NBN to a cost-benefit analysis some economists say that may be impossible. Last week, a report by Australian economic research group Access Economics stated there was insufficient data to estimate the benefits of a high-speed broadband network.

Still, Minchin said in a statement this morning, "This [OECD] advice makes a mockery of Communications Minister Stephen Conroy's naive and arrogant dismissal of the need for a thorough cost-benefit analysis."

The report did, however, call for governments to make information about broadband proposals available to the public, said Minchin. A recommendation he said ran counter to Conroy's refusal to make public the ACCC and expert panel reports into the first terminated $4.7 billion NBN tender process, and a factor that is likely to be a sticking point should the government require support in the Senate from key independents when it attempts to pass legislation for the NBN roll-out.

Topics: NBN, Broadband, Government AU

Liam Tung

About Liam Tung

Liam Tung is an Australian business technology journalist living a few too many Swedish miles north of Stockholm for his liking. He gained a bachelors degree in economics and arts (cultural studies) at Sydney's Macquarie University, but hacked (without Norse or malicious code for that matter) his way into a career as an enterprise tech, security and telecommunications journalist with ZDNet Australia. These days Liam is a full time freelance technology journalist who writes for several publications.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Universal access more important than speed

    I've argued for some years (including posts on ZDnet of over two years ago) that what affects productivity (on a national basis) is near-universal connectivity to the net, not the speed of such services per se. That way, we can stop producing paper-based directories; utilities and government will know you can communicate with people via email not snailmail; everyone can research any product, and access any electronic marketplace they like; etc.

    In this respect, I was chatting to former Deputy PM John Anderson just a couple of weeks ago, saying that the Coalition's rural access (satellite) subsidy was the most effective 'broadband' strategy yet implemented by either side of politics. The reason is that the subsidy (roughly the cost of original installation of a dish plus satellite modem) just sits there, for people to utilise if they so desire. So, rather than taking wires up to the front gate of a house that does not want technology, with the satellite subsidy, getting broadband presents nil upfront costs, for anyone prepared to pay the c$45/month usage fees of a (low-speed) satellite broadband link. At around 256-512kbps (ie max download of 50k Bytes per second) it is not blindingly fast, but it gives adequate connection for all those for whom dial-up over long/bad lines was just too slow... as that was not 52kbps but practically only 19kbps or just 2k Bytes per second. So a small one-time subsidy gives the most remote users a 25x speed improvement, and permanent connectivity. Such users can monitor weather (BOM site), news, rural markets (Landmark etc) and get email and downloads. They "FEEL" connected to the rest of the world. So, how much more would their productivity go up if you upped the speed by a further 10x or even 20x...

    I suggest that most of the gains have already been made. Interactive video calls are already possible with Skype on such connections.. and the people move jerkily, due to satellite latency, but would such people produce more wool, beef or minerals if the people on the screen moved less jerkily?

    The Japan and Korea experience is that the biggest benefactors of higher bandwidth is interactive gaming and stored-video watching. As I noted two years ago, there is a good argument that such activity is contrary to national productivity (people get absorbed and fail to contribute).

    Yes, I understand that there will be new apps - you'll have a resistance surface that makes you 'feel' like you're pushing on something on the other side of the web-link. But we can already turn on or turn off whatever valves etc we might have need to effect at a distance, and monitor results with remote cameras.... so the sensation will be more life-like, but it will not be associated with an improvement in productivity.

    One big issue is the term "highway" (the term 'super-highway' seems to have been dropped), as if building this NBN will be like the US build of its major highways as an employment-boosting project of the 1930s. Real highways do generate a lot of employment. But an NBN is really more like manner-from-heaven for offshore companies like Cisco. The problem is that vehicle highways need some maintenance, but basically stay productive. The NBN is technology, so may be more like buying a PC - you do so knowing at the outset that it will be out-of-date in just a few years, and become absolutely worthless at 5x the speed of a car's depreciation, and 20x the speed of a house or concrete roadway's physical deterioration. Moreover, Cisco charges a premium for its routing gear, whereas that technology will become commonplace over the next decade and prices will continue to drop at a fast rate.

    Indeed, the dual hybrid-fibre roll-out of just ten years ago (Telstra+Optus) is now so dated, that this plays no significant role in the newly-proposed NBN. That should stand as a warning. Optus cannot have made any realistic return on the $5b cost of that duplicate set of optical cable wiring serving just the s
  • ... Continuation of above post

    Sorry, above long post was truncated. It continued thus:
    Optus cannot have made any realistic return on the $5b cost of that duplicate set of optical cable wiring serving just the same medium-density suburbs as Telstra decided to also cable (while leaving vast swathes unserviced).

    And finally, no mater what set of technology is used, much of the country (weighted by square kilometres rather than population) will only be servicable by satellite, so we need the existing approach to rural and remote... and that is working fine.

    I'm all for new technology, but I too would love to see some genuine cost/benefit figures. I'm sure that having new not-Telstra-controlled inter-city high-bandwidth connections would be well-justified. Further, rolling out that access within the few tens of square kilometres of the major CBDs would also be well-justified. Connecting in major regional centres would be marginal (and compete with Telstra's existing bandwidth to such centres). And passing 95% of houses would prove very wasteful. The final 3-5% can only be done with satellite (as it should be - due to the cheap nature of adding extra remote users with that technology).

    The current error is to mix up the bits that have high return (opening the business bandwidth market up to competition post-Sol) versus the bits that must lose money (passing 95% of houses of which those interested already have a suitable connection). We should not do silly things just because they were promised in a bid-fest before a national election!
  • NBN is about SPIN

    OH my God. This is hight time where Krudd and Conjob get shown up for the spin they have been spinning.

    The only way out of this political mess and retain their election promises is to finally give some concessions to Telstra.

    If I was Thodey I would be prepared to walk away from this NBN pipedream unless this over regulated market id freed up from regulation allowing investment!!
  • And Pete Costello agrees...

    Shortly after submitting the above post, and a similar one on 'Telstra loses HFC access battle' dealing with how Sol's actions have forced the government into working out how to 'go around' Telstra re NBN... I've just now noticed that Peter Costello has written on the topic in today's SMH, of which I cite just a small extract (for reasons of copyright):

    "The only victim here is Telstra's share price
    SMH May 27 2009
    Peter Costello

    There are plenty of reasons to be critical of Sol Trujillo's performance as chief executive of Telstra...

    When he started, the Telstra share price was about $5. After four years it is about 40 per cent lower and hovering a little above $3...

    Burgess waged unrelenting war against the regulatory regime which allowed competition to Telstra on fair and reasonable terms as determined by the Australian Competition and Consumer Commission...

    It tried to intimidate and bully the regulator...

    There is no advantage in replacing a public monopoly with a private one. Competition meant withdrawing the privileges Telstra enjoyed as a government entity so other companies could compete on a fair basis. Telstra management liked the privatisation but they were very hostile to removing the monopoly privileges...

    The upshot is that when Labor's tender failed, the Rudd Government decided to commit taxpayer funds to a fibre network with no business plan, no guarantee of commercial return, no due diligence - with no process - just to save face on the collapsed tender. It is a gamble of unprecedented proportion. Unless the Government listens to some reason about this proposal it could be one of Australia's great commercial disasters...
    [end of partial extract]
  • Getting rid of regulation is NOT the answer

    Anon says what's needed is "this over-regulated market is freed up from regulation".

    That won't fix anything. What is needed is a clear formula for regulation. Giving a monopoly right to any private company is crazy, as Peter Costello explained. There are already other telcos in Australia, and we are not about to force them to leave.... Sooo, we must set up a regime for access and pricing.

    The most successful utility regulators overseas use return on assets employed to determine pricing to deliver market returns on monopoly aspects of a service. That way, Telstra can charge what they like to a retail customer, but need to comply with pricing for competitors which returns them a good return, but not a 'monopoly rent'.

    If Telstra fell over and said that post-Sol it will accept regulated prices for just those components which are effective monopolies, it would find life far easier in Australia... and Australia would worry far less about the elephant in the room. The alternative is a complete split, with Telstra Long Lines (say) able to play a significant role in the NBN, as it would supply a wholesale service to ALL telcos, including Telstra Retail.
  • A concrete example calc of productivity gains

    My first post was 'Universal access more important than speed', noting that access to electronic information was more important for national productivity than the speed per se. But I never explained with an example.

    Let's say a typical web-page (on existing broadband technology) takes 3-4 seconds to load., but takes 30-60 seconds for the human user to read and process the information contained. Now let's examine the effect of a ten-fold increase in internet speed. For conservatism we ought use the 'worst' figures from the above average range (4 seconds to load yet only 30 seconds to read). With the 10x speed increase, the page will now load in 0.3 seconds. But what happens in terms of time improvements for the 'overall task'. The time to 'load and read' will only reduce from 33 seconds to 30.3 seconds.
    A 900% improvement in internet speed resulted in only an 8% improvement in productivity WHILE actively using the internet. So let's say that the average worker spends even a full hour just reading web-based information each working day. That means a 900% improvement in internet speed will give an overall improvement to worker productivity of just 1%... but in fact a whole lot of people do NOT spend much time on the internet for their work, so the one-percent figure is a gross over-estimation.

    Yes, it would improve the game-satisfaction of on-line interactive gamers by 2x (say) and allow movie addicts to watch 3x as many movies.. but both of those activities are negatively correlated with national productivity. It could also improve interactive remote learning, but I suspect the big improvements to be made in this space will be in more complete on-line course content and software-enabled student-teacher interactivity, rather than IP packet speed.
  • Another Initiative Would Deliver Greater Productivity

    So, if it is not cost-benefit wise for the Feds to spend more than 10% of the alleged $43b for the NBN, what ought they do?

    Krudd could simply say 'Ah, I'm now informed that a similar E-initiative has a far greater payback, so we're doing that instead of that crazy pre-election bid-fest both parties announced re the NBN."

    Giving every citizen their own unique email address and managing that for them, would produce far higher productivity gains. The feds ought let a contract for an IT contractor to develop an entirely open-format XML-based database (database indexing-system-independent) of all the fields 'reasonably needed' to be maintained about a citizen or registered company or non-profit. At a minimum it would have a unique email address for each citizen. That might be a eleven-digit number, allowing 999 million citizens in 9-digits and incorporating two check digits. The 'formal' email address would be those digits followed by "@au.gov.au" or "@people.au" etc. However, the system would allow users to log-on directly and edit their own settings, allowing this email address to also be forwarded to their 'favourite' normal email address, as specified from time to time by them. However, you could ALWAYS log on and see exactly what was sent to you by government or utilities etc on the actual web site (or via any POP-enabled email app). Storing your official email history 'virtually forever' for free is still 10x cheaper than sending you just one government notification by snail-mail.

    Anyway, if you change address, surname etc, you only need edit that info in one spot, and ALL government agencies would be automatically updated. There is a need to ensure that it does not involve the worst bits of the original Australia Card, but that could be managed.

    Arguably, to get REALLY SIGNIFICANT productivity gains, in a Phase 2, you insist that this database is THE database for any government department to keep information on you. Firstly, this lets the actual user(citizen) see what the government has on its records for themselves, with an ability to provide on-line feedback to correct some entry or other. All government departments would then be forced to keep their information 'visible', overcoming much of the FOI problems of the past. With XML technology, even local councils can add records, noting that your name is 'associated' with a list of property ID numbers and a rate-payer number.

    The health-related records would be in a specially-sealed section, accessible only by Medicare and doctors who abided by the code of conduct to look-up only those patients currently under their care (or face being struck off).

    Still far longer-term, one could insist on a truly whole-of-government approach, whereby each government department (but for approved exemptions for ATO, security etc) HAS to have all of its information about you ONLY on the site where you can see what they hold.

    Separate QANGO instrumentalities such as rural lands protection boards, could then have their database just a subset of the total national database. Only one good datacentre is needed, not hundreds. Backup of all data could be achieved nationally by a couple of individuals. Separate little databases, such as the 'Do Not Call' register would be included, and be self-service, rather than having their own separate administration/management.

    But if you were open to contact, you could even add 'external visibility' for your phone number, street address and preferred email address, so if you moved, your relatives, former friends etc could re-find you. However, if you wanted complete privacy, that too would be settable. But if you had supplied your national email address to your banks, utility suppliers, telco suppliers, etc, when you did move physically, you might tick a box to say just notify those corporate contacts to whom you have already given your email address... and all of your relocation details are promulgated, even with banks etc.

    The fact t
  • ... Continuation of above post

    The fact that it could optionally have your promoted web-site address (url) or social networking preferred link, means it would fully replace the current below-par White Pages functionality... as well as removing the role of maintaining the national register away from a private company (which claims copyright over your address details).

    This approach could be both friendly and a very great addition to national productivity.
  • The Minchin gang (Abbott and the Wacko bolgshop senator from South Australia) are halfwits from the far out loony right, it's about time the Liberals got rid of this senile old fool.
    There is no such thing as a cost benefit analysis, it's an exercise in hypotheticals invented by politicians.
    In the end any cost benefit analysis is a waste of time as the politicians that don't like the results will just argue about the methodology and the end results.
    The full cost of the NBN is unknown it could be more or less expensive than currently forecast.
    The full benefits are largely unknown as the network is not yet implemented and applications not yet fully developed. In the days of dialup plenty of people argued that broadband was not needed or relevant they were wrong.
    As a general rule people/"experts" always underestimate usage of new communications technology.
    The only certainties are the NBN will vastly improve telecommunications particularly for rural and regional areas, and will allow more broadcasting channels for smaller operators as the present cable monopolies are broken down, and will allow for much better video telephony.
    Kevin Cobley